Periodic Reporting for period 3 - ARTISTIC (Advanced and Reusable Theory for the In Silico-optimization of composite electrode fabrication processes for rechargeable battery Technologies with Innovative Chemistries)
Période du rapport: 2021-04-01 au 2022-09-30
The development of tools for the accelerated optimization of the manufacturing process of secondary batteries (such as Lithium Ion Batteries) is crucially needed in order to reduce the associated CO2 fingerprint and cost while improving the electrochemical performance of the manufactured batteries. The aim of the ARTISTIC project is to develop and to demonstrate a novel theoretical framework devoted to rationalizing the formulation of composite electrodes containing next-generation material chemistries for high energy density secondary batteries. The framework is being established through the combination of discrete particle models, continuum mathematical models and Artificial intelligence/Machine Learning models within a multiscale computational workflow integrating the individual models and mimicking the different steps along the electrode manufacturing process, including slurry preparation, drying, calendering and electrolyte filling. Strongly complemented by dedicated in house experimental characterizations which are devoted to its validation, the goal of this framework is to provide insights about the impacts of material properties and manufacturing process parameters on the electrode mesostructures and their corresponding correlation to the resulting electrochemical performance. It targets self-organization mechanisms of material mixtures in slurries by considering the interactions between the active and conductive materials, solvent, binders and dispersants and the relationship between the materials properties such as surface chemistry and wettability. Optimal electrode formulation, manufacturing process and the arising electrode mesostructure can then be achieved. Additionally, the framework will be integrated into an online and open access infrastructure, allowing predictive direct and reverse engineering for optimized electrode designs to reach high electrochemical performances. Through the demonstration of a multidisciplinary, flexible and transferable framework, this project has tremendous potential to provide insights leading to proposals of new and highly efficient industrial techniques for the manufacturing of cheaper and reliable next-generation secondary battery electrodes for a wide spectrum of applications, including Electric Transportation.
Until now, we have carried out several major achievements, namely the development/demonstration:
(i) of innovative computational procedures based on Coarse-Grained Molecular Dynamics (CGMD) to consistently model and validate 3D Li-ion battery (LIB) electrode slurries and to accelerate their parametrization;
(ii) of novel computational procedures to mimic the electrode calendering step performed in LIB manufacturing (Discrete Element Method -DEM-);
(iii) of the INNOV algorithm, allowing to mesh 3D electrode mesostructures and to input them in continuum models able to assess their electrochemical performance;
(iv) of a novel 3D-resolved Lattice Boltzmann model to simulate the electrolyte filling step;
(v) of a novel 4D (x, y, z, time) model simulating the discharge process of the electrodes;
(vi) of a novel 4D model simulating the impedance spectroscopy response of the electrodes;
(vii) of in silico workflows coupling sequentially the different simulation techniques mentioned above (e.g. using the electrodes predicted by CGMD as inputs of the 4D performance model). We have also developed a methodology to track uncertainty propagation while coupling these models;
(viii) of a novel batch of machine learning (ML) models aiming to predict the effect of LIB electrode manufacturing on their associated properties;
(ix) of a consequent in house experimental database for the parameterization and validation of the models above;
(x) of a first version of the online manufacturing simulator (data explorer) becoming accessible to the community upon login creation (https://www.erc-artistic.eu/computational-portal/).
(xi) of the integration of ARTISTIC results (e.g. calculated battery electrode structures) in Virtual Reality (VR) serious games we are using for teaching about battery technologies at the University level and in science popularization events.
(i) of innovative computational procedures based on Coarse-Grained Molecular Dynamics (CGMD) to consistently model and validate 3D Li-ion battery (LIB) electrode slurries and to accelerate their parametrization;
(ii) of novel computational procedures to mimic the electrode calendering step performed in LIB manufacturing (Discrete Element Method -DEM-);
(iii) of the INNOV algorithm, allowing to mesh 3D electrode mesostructures and to input them in continuum models able to assess their electrochemical performance;
(iv) of a novel 3D-resolved Lattice Boltzmann model to simulate the electrolyte filling step;
(v) of a novel 4D (x, y, z, time) model simulating the discharge process of the electrodes;
(vi) of a novel 4D model simulating the impedance spectroscopy response of the electrodes;
(vii) of in silico workflows coupling sequentially the different simulation techniques mentioned above (e.g. using the electrodes predicted by CGMD as inputs of the 4D performance model). We have also developed a methodology to track uncertainty propagation while coupling these models;
(viii) of a novel batch of machine learning (ML) models aiming to predict the effect of LIB electrode manufacturing on their associated properties;
(ix) of a consequent in house experimental database for the parameterization and validation of the models above;
(x) of a first version of the online manufacturing simulator (data explorer) becoming accessible to the community upon login creation (https://www.erc-artistic.eu/computational-portal/).
(xi) of the integration of ARTISTIC results (e.g. calculated battery electrode structures) in Virtual Reality (VR) serious games we are using for teaching about battery technologies at the University level and in science popularization events.
We believe that the achievements mentioned above advance very significantly the LIB electrode manufacturing process rationalization and the computational methodologies aiming to simulate it:
- achievement (i) introduces a new methodology which eases the use of CGMD for the simulation of multiple types of slurry chemistries;
- (ii) introduces an experimentally validated DEM model allowing to describe explicitly the mechanical interactions between active material particles and carbon-binder domain (CBD) (more realistic than considering only active material as in the few previously reported DEM models in litterature);
- (iii) is a novel generic code which can mesh any kind of complex structure with any kind of number of materials types, which allows accounting for different physics for the active material and CBD in the electrochemical performance models in achievements (v) and (vi);
- (iv) introduced a new model able to predict, with 3D resolution, how liquid electrolyte penetrates in the predicted electrode mesostructures;
- (v) and (vi) allowed for the first time simulating the electrochemical response of the electrodes by accounting by the explicit location of both active material particles and CBD, in stark contrast to previously reported litterature where active material and CBD were considered as a single solid phase with the same physical properties (e.g. electronic conductivity);
- (vii) constitutes an inherent novelty of this project (the combination within a single computational workflow of discrete and continuum models, in order to establish the link between chemistry/formulation/manufacturing process and electrochemical performance);
- (viii) introduced the first demonstration of the application of ML to discover correlations between LIB electrodes formulation/manufacturing parameters paving the way to the optimization of the manufacturing process;
- (ix) constitutes very important data ressources for manufacturing/electrochemical models parameterization and validation;
- There is no equivalent of the achievement (x) as far as we know. It aims allowing users to explore the data produced in the project in interactive way;
- Regarding (xi), it is the first time that VR serious games are developed/demonstrated for teaching battery concepts. Through strong collaboration with ergonomers/cognitive scientists, we have shown that VR helps at understanding complex properties and operation principles of battery electrodes. These VR serious games are also efficient at showcasing the ARTISTIC results in front of students and the general public. We have used these tools in University lectures, and we have performed demonstrations in conferences and science popularization events (e.g. Pint of Science, Science is Wonderful, etc.).
The expected results until the end of the ARTISTIC project include the study of other electrode chemistries, deeper insights of the manufacturing steps, the demonstration of the optimization of electrode architectures and the release of the web application allowing to perform manufacturing/electrochemistry simulations and optimizations online.
- achievement (i) introduces a new methodology which eases the use of CGMD for the simulation of multiple types of slurry chemistries;
- (ii) introduces an experimentally validated DEM model allowing to describe explicitly the mechanical interactions between active material particles and carbon-binder domain (CBD) (more realistic than considering only active material as in the few previously reported DEM models in litterature);
- (iii) is a novel generic code which can mesh any kind of complex structure with any kind of number of materials types, which allows accounting for different physics for the active material and CBD in the electrochemical performance models in achievements (v) and (vi);
- (iv) introduced a new model able to predict, with 3D resolution, how liquid electrolyte penetrates in the predicted electrode mesostructures;
- (v) and (vi) allowed for the first time simulating the electrochemical response of the electrodes by accounting by the explicit location of both active material particles and CBD, in stark contrast to previously reported litterature where active material and CBD were considered as a single solid phase with the same physical properties (e.g. electronic conductivity);
- (vii) constitutes an inherent novelty of this project (the combination within a single computational workflow of discrete and continuum models, in order to establish the link between chemistry/formulation/manufacturing process and electrochemical performance);
- (viii) introduced the first demonstration of the application of ML to discover correlations between LIB electrodes formulation/manufacturing parameters paving the way to the optimization of the manufacturing process;
- (ix) constitutes very important data ressources for manufacturing/electrochemical models parameterization and validation;
- There is no equivalent of the achievement (x) as far as we know. It aims allowing users to explore the data produced in the project in interactive way;
- Regarding (xi), it is the first time that VR serious games are developed/demonstrated for teaching battery concepts. Through strong collaboration with ergonomers/cognitive scientists, we have shown that VR helps at understanding complex properties and operation principles of battery electrodes. These VR serious games are also efficient at showcasing the ARTISTIC results in front of students and the general public. We have used these tools in University lectures, and we have performed demonstrations in conferences and science popularization events (e.g. Pint of Science, Science is Wonderful, etc.).
The expected results until the end of the ARTISTIC project include the study of other electrode chemistries, deeper insights of the manufacturing steps, the demonstration of the optimization of electrode architectures and the release of the web application allowing to perform manufacturing/electrochemistry simulations and optimizations online.